ethical questions, such as whether
algorithms plant viruses in someone
else’s device. This is clearly illegal and
unethical. There are cases where a social bot might ethically violate the law,
such as civil disobedience for a cause
the creator considers just. However,
civil disobedience is only ethical in
very rare cases in constitutional de-mocracies where legal recourse for
unjust laws pervade. 6 Cases where a
law may be broken that are not unethical require justification—compelling
arguments that appeal to moral standards of the majority. 6 Only in such
rare cases may illegal acts be seen as
moral and therefore ethical. 6 Thus we
ask “Is the illegal act justifiable?” Acts
that are not suitably justifiable (that
is, do not appeal to the morality of the
majority) are unethical. Swiss authorities did not file charges against the
Random Darknet Shopper developers.p
They argued that social bots can buy
illegal narcotics over the Internet for
the purpose of artq and that “ecstasy
in this presentation was safe.” The
behavior was not unethical because it
was justified according to the pervading morality of the community.
If a social bot’s behavior does not
break any laws, next evaluate for truth-fulness: “Is any deception involved?” Social bots may act deceitfully. For example, they can misrepresent themselves
as human beings2 or spread untruthful information (such as fake news).
Deceiving acts communicate false or
erroneous assertions, violating the
prima facie duty of fidelity. Social bots
should always act truthfully. 3 However,
deceitful acts can be justifiable if the
duty of fidelity is superseded by a high-er-order duty, such as beneficence.r
Deceptive, satirical actions may not
be unethical since they elicit pleasure,
improving the life of others. Consider
Big Data Batmans as an illustration.
p By “developer” we are referring to either the
organization or management of the organiza-
tion or the software developer involved in the
creation of the social bot.
r Beneficence is the duty to bring virtue, knowl-
edge or pleasure to others; other duties, ac-
cording to Ross 1930, include non-malefi-
cence, self-improvement, justice, gratitude,
reparation (see Mason et al. 7, p. 132–133).
there are shades of gray that are difficult to judge.
For example, Tay,l a social bot created by Microsoft to conduct research
on conversational understanding,
went from “humans are super cool”
to “Hitler was right I hate the Jews”
in less than 24 hours on Twitter due
to malicious humans interacting
with the social bot.m In another case,
a social bot tweeted “I seriously want
to kill people” from randomly generated sentences during a fashion
convention in Amsterdam.n Clearly
such inadvertent comments violate
our sensibilities and are distasteful, but are they unethical? Perhaps,
but by what standard do we judge?
Some social bots do more than just
comment—clearly those that steal
information and other misdeeds
are engaging in unethical activity,
but, again, it is not always so clear.
For instance, the Random Darknet
Shopper—a social bot coded to explore the dark Web in the name of
art—inadvertently purchased 10 Ecstasy pills (an illegal narcotic) and a
counterfeit passport.o So a law was
broken, but was this unethical behavior? We developed a procedure,
which we describe next, to help answer such questions.
Bot Ethics: A Procedure to Evaluate
the Ethics of Social Bot Activity
Ethics in philosophy dates back thousands of years, and this Viewpoint column cannot do justice to the entire
field. However, because of the increasing prominence of social bots and their
potential for malicious activity, ethical
judgment about their activity is necessary. The best way to guide ethical
conduct in a community is to provide a
procedure for reflection and discourse. 5
The procedure we created is called “Bot
Ethics” (see the figure here) and it focuses on the behavior of social bots with
respect to law, deception, and norms.
Many laws are developed from ethical
principles. 6 Even when a law may be
flawed, it is typically the ethical course
of action to follow that law. 9 Therefore
a natural first question is: “Does the action of the social bot break the law?” The
objective is to assess straightforward
Bot Ethics: How to determine whether social bot actions are unethical.
1. Break Law?
If Evil, Less
Social bots have been
reported to behave
badly in a variety
of ways across